Matching 2D and 3D Articulated Shapes using Eccentricity

Adrian Ion (1) , Nicole M. Artner (1) , Gabriel Peyré (2) , Walter G. Kropatsch (1) , Laurent D. Cohen (2)

1 PRIT - Pattern Recognition and Image Processing Group
2 CEREMADE - CEntre de REcherches en MAthématiques de la DEcision

Shape matching should be invariant to the typical intra-class deformations present in nature. The majority of shape descriptors are quite complex and not invariant to the deformation or articulation of object parts. Geodesic distances computed over a 2D or 3D shape are articulation insensitive. The eccentricity transform considers the length of the longest geodesics. It is robust with respect to Salt and Pepper noise, and minor segmentation errors, and is stable in the presence of holes. We present a method for 2D and 3D shape matching based on the eccentricity transform. Eccentricity histograms make up descriptors insensitive to rotation, scaling, and articulation. The descriptor is highly compact and the method is straight-forward. Experimental results on established 2D and 3D benchmarks show results comparable to more complex state of the art methods. Properties and results are discussed in detail.

Mots clés: Eccentricity Transform; Shape Matching; Articulation; Geodesic Distance

Domaines: Traitement du signal et de l'image [eess.SP]; Traitement du signal et de l'image [eess.SP]

Adrian Ion, Nicole M. Artner, Gabriel Peyré, Walter G. Kropatsch, Laurent D. Cohen. Matching 2D and 3D Articulated Shapes using Eccentricity. 2008. ⟨hal-00365019v1⟩